- Email: [email protected]
Results of a protocol of transfusion threshold and surgical technique on transfusion requirements in burn patients
Burns 31 (2005) 558–561 www.elsevier.com/locate/burns
Results of a protocol of transfusion threshold and surgical technique on transfusion requirements in burn patients Michael S. O’Mara a,b, Fernando Hayetian c, Harvey Slater c, I. William Goldfarb c, Eric Tolchin c, Philip F. Caushaj c,* a
Shriner’s Hospital for Children of Northern California, Department of Burn Surgery, Sacramento, CA, USA b The University of California Davis Medical Center, Department of Surgery, Sacramento, CA, USA c The Western Pennsylvania Hospital, Department of Surgery, Burn Trauma Unit, Temple University School of Medicine Clinical Campus, 4800 Friendship Avenue, Pittsburgh, PA 15224, USA Accepted 6 January 2005
Abstract Introduction: Blood loss and high rates of transfusion in burn centers remains an area of ongoing concern. Blood use brings the risk of infection, adverse reaction, and immunosuppression. Methods: A protocol to reduce blood loss and blood use was implemented. Analysis included 3-year periods before and after institution of the protocol. All patients were transfused for a hemoglobin below 8.0 gm/dL. Results: Operations per admission did not change during the two time periods (0.78 in each). Overall units transfused per operation decreased from 1.56 0.06 to 1.25 0.14 units after instituting the protocol ( p < 0.05). Also, units transfused per admission decreased from 1.21 0.15 to 0.96 0.06 units of blood ( p < 0.05). This was noticed particularly in burns of less than 20% surface area, declining from 386 to 46 units after protocol institution, from 0.37 to 0.04 units per admission, and from 0.79 to 0.08 units per operation in this group of smallest burns. There was no change noted in the larger burns. Conclusions: This study suggests that a defined protocol of hemostasis, technique, and transfusion trigger should be implemented in the process of burn excision and grafting. This will help especially those patients with the smallest burns, essentially eliminating transfusion need in that group. # 2005 Elsevier Ltd and ISBI. All rights reserved. Keywords: Blood loss; Burn centers; Transfusion
1. Introduction Blood loss during excision and grafting of thermal injuries has long been of concern. Transfusion requirements have remained significant, with exsanguination sometimes life-threatening. Although replacement of blood loss is simple, the risk associated with transfusion is not insignificant. Blood products are related to a direct risk of infection [1], as well as an increased susceptibility to infection [2]. Also, transfusion has been implicated in the reduction of immunocompetence [3,4]. Consideration of these factors * Corresponding author. Tel.: +1 412 578 4024 (O); fax: +1 412 578 1434. E-mail address: [email protected] (P.F. Caushaj). 0305-4179/$30.00 # 2005 Elsevier Ltd and ISBI. All rights reserved. doi:10.1016/j.burns.2005.01.006
has led to an attempt to decrease transfusion requirements in burn patients, as well as the general surgical population [5]. A multitude of hemostatic techniques have been developed in recent years. Traditional methods of pressure and tourniquets are now being combined with chemotherapeutic agents [6]. Thrombin, epinephrine [7], and fibrin [8] are the most common modalities that have been used to minimize blood loss. This becomes vital as tangential excision is widely implemented, with the associated potential for increased blood loss. The traditional method of excision to fascia decreases operative blood losses, but is problematic for the extremities, as the sequelae of poor functional outcome are increased [9]. The benefits of minimizing blood loss become more apparent as strategies
M.S. O’Mara et al. / Burns 31 (2005) 558–561
involving a lowered threshold for transfusion are implemented, decreasing the number of patients who qualify for blood. Considering the advances in techniques to minimize blood loss and transfusion requirements, we retrospectively analyzed our institution’s transfusion practices prior to and following institution of a protocol to decrease intraoperative losses and de-liberalize transfusion threshold. In particular, we had noted that patients with smaller burns, those less than 20% total body surface area, rarely needed transfusion, and thus the risk seemed to have been eliminated. We sought to determine if a reduction in blood usage had occurred.
2. Material and methods Two 3-year time periods were analyzed, before and after implementation of our intraoperative protocol, to reduce blood loss. In the earlier period (1993–1995), methods of excision and grafting were more variable. In general, excision and harvesting was done initially, with coverage of sites with thrombin soaked and/or epinephrine soaked gauze. Pressure was applied to all sites for a minimum of 10– 15 min. Electrocautery was used to coagulate individual punctate hemorrhage, and pressure reapplied as needed. Hemostasis was achieved before grafting. Excision was done by a mixture of tangential excision with use of excision to fascia for extensive, deep burns of the thorax. In the later period (1997–1999), a protocol was followed, with modification based upon the attending surgeon’s clinical judgment. Extremities were excised tangentially under tourniquet control. The thorax was excised to fascia for burns of greater than 10% contiguous surface area involvement. Tangential excision was used for small, localized regions of the thorax (<5% contiguous area) and for female breast involvement. Whenever possible the nipple complex was preserved. Excision was done in an expeditious fashion, with two teams working in concert when more than one body region required excision. OmidermTM (Omiderm Ltd., commercially available) synthetic film sprayed with thrombin solution (GenTrac Inc., commercially available) was placed over all areas of excision and epinephrine-soaked gauze (1:1000 injectable, 1 mg, mixed in 1 L normal saline, American Regent Laboratories, commercially available) used to cover this. Pressure was then applied over the wound. While hemostasis was achieved, grafts were harvested. Over the split thickness donor site omidermTM, thrombin, and epinephrine soaked gauze were also applied. After 15 min of pressure, gauze and omidermTM were removed and wounds examined. Rare electrocautery was needed. Larger areas of diffuse capillary bleeding were re-covered until hemostasis was achieved. Donor skin was then meshed and grafted, and dressings applied. Postoperative bleeding was controlled with local electrocautery or direct suture ligation in the burn unit. The same two surgeons performed all procedures during both
559
time periods (IWG, HS). All procedures were performed at one hospital with patients all in the same burn unit (the Western Pennsylvania Hospital Burn Trauma Unit). During both periods, patients were evaluated for transfusion in the same manner. All patients were transfused for a hemoglobin below 8.0 gm/dL. Patients with a significant cardiac history and who were critically ill were transfused to a hemoglobin of 10.0 gm/dL. These patients were not considered separately. Patients were not given prophylactic transfusions in anticipation of blood loss if their hemoglobin concentration met these criteria. Intraoperative transfusion was reserved for patients who were hypotensive or tachycardic or were clinically judged to have excessive blood loss. Number of admissions, mean body surface area of burns, number of operations, area grafted per patient, units of blood transfused, units transfused per admission, and units transfused per operation were compared between the two time periods. All patients admitted with burn injuries were included in the evaluation. Criteria excluded only those patients admitted to the burn trauma unit who were not actually burn injuries (i.e. coumadin skin necrosis, toxic epidermal necrolysis). All thermal, chemical, and electrical burns were included whether or not operative interventions or transfusions were required. Patients refusing blood transfusion for religious reasons or otherwise were included. All operations were considered together, including those for re-grafting. Patient groups were further subdivided by burn size. The same evaluation was carried out, comparing small burns (0– 20%), moderate burns (20–40%), and large burns (>40%). Statistical analysis by two-tailed t-test was done using GraphPad InStat Version 3.01 for Windows 95/NT, GraphPad Software, San Diego, CA, USA (www.graphpad.com).
3. Results Total values were compared between the two time periods. Data are shown in Table 1. The later protocol group appeared to have smaller burns in all 11.2 14.5% versus 13.0 15.7% in the earlier group, and this was significant ( p = 0.01). Most of this difference appeared to be among the patients with smaller burns (see below). There was no change in number of operations per admission, 0.8 0.07 in the early group and 0.8 0.1 in the later, p = 1.0. No difference was noted in area grafted per patient, 995 cm2 in the early group and 1010 cm2 in the later, p = 0.9. There was a significant decrease in transfusions per admission and per operation. Transfusions per admission declined from 1.2 0.16 to 0.96 0.06 units ( p < 0.0001) and transfusions per operation from 1.6 0.14 to 1.3 0.14 units ( p < 0.0001). In the two groups of larger burns (20–40% and >40% surface area burns), there did not appear to be any change in transfusion requirements between the two time periods.
Total—all patients, all sizes of burn; small—patients with burns <20% surface area; mod—patients with burns 20–40% area; large—patient with burns >40% area. * p < 0.01, otherwise p > 0.05.
52 212 505 63.7 19.0 2685 2180 4.1 5.3 9.7 11.5 2.4 2.7 98 198 512 27.5 5.4 2060 1050 2.0 1.3 * 5.2 1.7 2.6 .67 111 215 542 28.7 5.5 1980 1120 1.3 1.2 * 4.9 1.6 2.5 .56 1129 567 46 6.6 5.3* 850 650 0.50 0.15* 0.04 0.01* 0.08 0.01* 1027 490 386 7.3 5.2* 805 605 0.48 0.10* 0.37 0.08* 0.79 0.12* 1271 981 1219 11.2 14.5* 1010 3820 0.78 0.11 0.96 0.06* 1.3 .14 * 1190 917 1433 13.0 15.7* 995 3870 0.78 0.07 1.2 .16 * 1.6 .14 * Admissions Operations Transfusion units Area of burn (%) Grafted area (cm2) Operation/admission Transfusion/admission Transfusion/operation
Large
Early Late
Mod
Early Late
Small
Early Early
Late Total
Table 1 Admission, operation, and transfusion data
44 216 661 64.5 18.8 2790 2040 4.9 5.9 15.0 14.9 3.1 5.3
M.S. O’Mara et al. / Burns 31 (2005) 558–561
Late
560
Burn size was comparable between the two time periods as well, as was the area grafted. The later group of moderate burns had more operations per admission, 1.3 1.2 versus 2.0 1.3 ( p < 0.001), but no change in transfusion need (Table 1). There was a trend in the large burns toward greater transfusions per admission and per operation, although this was not significant. The earlier group required 9.7 11.5 versus 15.0 14.9 units per admission in the later group, p = 0.051. And per operation, the earlier group was transfused 2.4 2.7 units and the later 3.1 5.3 units, p = 0.1 (Table 1). Consideration of the small burns revealed the greatest differences. In this segment, the later group had smaller overall burns, 6.6 5.3% versus 7.3 5.2% ( p = 0.009). Despite this, the later group had more surgical interventions, 0.50 0.15 versus 0.48 0.10 operations per admission in the earlier group, p = 0.001. This did not translate to a larger area of grafting. With more operations, smaller burns, and the hemostatic protocol, the later group had a significant drop in blood requirements. Grossly, this was a drop from 386 to 46 units over the respective time periods. The later period required fewer units per admission, 0.04 0.01 from 0.37 0.08, p < 0.0001. This protocol group also received fewer transfusions per operation, 0.08 0.01 units from 0.79 0.12 units earlier, p < 0.0001 (Table 1).
4. Discussion There has been a move across the spectrum of surgical specialties to reduce the necessity for blood transfusions. The benefits of decreased risk of blood borne infections, transfusion reactions, and immunosuppression have been documented [1–5]. In burn patients, it is as important to limit blood losses as it is to limit transfusion volume. Risk to cardiac function and hemodynamics of the patient are evident with exsanguination during burn surgery, and thus techniques to decrease these losses can both improve patient intraoperative risk as well as decreasing the need for postoperative transfusion. The evidence for risk of blood-born transmission and increased risk of infection with transfusion has been elucidated. An increased risk of bacterial infection exists postoperatively in the transfused patient. It is unlikely that these infections stem from infected units, but rather from host modulation that increases risk of clinical infection. The cause and effect is unclear, but the mere act of giving blood is associated with a likelihood of contamination [2]. HIV has been estimated to be a risk in up to 1 of every 200,000 units transfused, while nearly 1% of blood recipients are at risk of contracting hepatitis [10]. Screening of blood products is improving, but the risk remains. Immunosuppression with transfusion has been documented. Early recurrence of colorectal cancer is associated with perioperative transfusion, as well as increased likelihood of
M.S. O’Mara et al. / Burns 31 (2005) 558–561
postoperative infection. Leukocyte depletion in transfused units appears to be associated with improved outcomes, and may suggest the mechanism of immunomodulation. As of yet the clinical significance of this has not become clear, and results are conflicting as to the cause and effect of changed outcomes [10]. In studies of the critically ill and of surgical patients, hemoglobin levels as low as 7.0 gm/dL have shown no increase in mortality [11,12]. This is true in the majority of patients, but in patients who are at increased risk for cardiac mortality and who are critically ill, a hemoglobin of 9.0 gm/ dL has been suggested as being more prudent for adequate oxygenation and perfusion [12]. Prophylactic transfusions have not been shown to be of particular value, and may put a patient at unnecessarily increased risk [12,13]. Current practices at burn centers reveals a broad array of transfusion triggers and guidelines are being used [14]. No standard care exists despite current literature implying the efficacy of minimizing transfusions to critically ill patients. Overall we saw a decrease in transfusion use with the change in technique of hemostatic management. The majority of the decline in units of blood used appears to have been in the patients with the smallest burns. It is this group of patients where the benefit is also best seen. In larger burns, there would be only a modest benefit from decreasing the number of units transfused per patient or per operation, as the risk of transfusion is still present. In the smaller burns, though, there is a notable advantage in that transfusion was successfully eliminated in the majority of patients who previously would have been given blood. In the earlier time period, nearly 1 in 3 patients admitted received a transfusion, while after the protocol was instituted, this dropped to less than 1 patient transfused in every 25. The benefit to these patients, the largest group of burn patients, is considerable. We no longer type our patients with small burns for possible transfusion, and have attempted to decrease our frequency of evaluation of hemoglobin concentrations. The trend in the larger burns toward a greater transfusion requirement is concerning. This occurred despite no real difference in burn size. One contribution to this could be a trend toward more frequent, smaller excisions and grafting. We speculate that our confidence in better hemostasis allowing these patients to tolerate more has spurred us to be more aggressive in our excision and grafting of areas of burn we would have treated more conservatively in the past. This increased excision and grafting could benefit patients with a better functional outcome. Other modalities of decreasing blood loss may be more appropriate than our current methods, such as subeschar infusion of epinephrine and fibrin modalities. This could also decrease the need for excision to fascia in larger burns. This continues to be a process in evolution. Multiple means of decreasing blood loss and corresponding transfusion requirements have been used in the burn population. Techniques have been shown to individually
561
decrease losses. More paramount recently has been the identification of a need to move to a lower transfusion trigger. On combining these chemotherapeutic, surgical, and transfusion protocol methods, it is possible to bring about a coordinated diminution of transfusion. In order to optimize functional results, a certain level of bleeding must be accepted. By focusing efforts, we have shown it is possible to decrease overall use of blood in our burn unit, and in particular to eliminate transfusion requirements in a large segment of the burn population. Patients with larger burns can be more aggressively debrided and grafted toward better functional outcomes. Further studies comparing like burn sizes with direct measure of blood loss and hemoglobin change may be indicated, but we suggest that this is only necessary to confirm what we see, which is an improvement by decreasing blood need with the efforts of a burn team to minimize loss. This benefit is not limited to one specific protocol, but rather is a result of having a defined set of techniques and parameters in place. Such efforts decrease the multifactoral risk to the patient, and will undoubtedly improve outcomes.
References [1] Graves TA, Cioffi WG, Mason AD, McManus WF, Pruitt Jr BA. Relationship of transfusion and infection in a burn population. J Trauma 1989;29:948–52. [2] Triulzi DJ, Blumberg N, Heal JM. Association of transfusion with postoperative bacterial infection. Crit Rev Clin Lab Sci 1990;28:95– 107. [3] Persijn CG, Cohen B, Lansbergen Q, vanRood JJ. Retrospective and prospective studies on the effect of blood transfusions in renal transplantation in the Netherlands. Transplantation 1979;28:396–401. [4] Tartter PI. The association of perioperative blood transfusion with colorectal cancer recurrence. Ann Surg 1992;216:633–8. [5] Mann R, Heimbach DM, Engrav LH, Foy H. Changes in transfusion practices in burn patients. J Trauma 1994;37:220–2. [6] Sawada Y, Yotsuyanagi T. A technique of haemostasis of the extremities after debridement of burn wounds. Burns 1992;18:412–5. [7] Snelling CFT. Shaw K. The effect of topical epinephrine hydrochloride in saline on blood loss following tangential excision of burn wounds. Plast Reconstr Surg 1983;72:830–4. [8] Ihara N, Suzuki K, Tanaka H, Nakamura Y, Tanabe J, Fukuda A, et al. Application of fibrin glue to burns. Burns 1984;10:396–404. [9] Achauer BM, Hernandez J, Parker A. Burn excision with intraoperative vasopressin. J Burn Care Rehabil 1989;10:375–8. [10] Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion medicine. NEJM 1999;340:438–47. [11] Carson JL, Duff A, Jerlin JA, Lawrence VA, Poses RM, Huber EC, et al. Perioperative blood transfusion and postoperative mortality. JAMA 1998;279:199–205. [12] Hebert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. NEJM 1999;340:409–17. [13] Valeri CR, Crowley JP, Loscalzo J. The red cell transfusion trigger: has a sin of commission now become a sin of omission? Transfusion 1998;38:602–10. [14] Palmieri TL, Greenhalgh DG. Blood transfusion in burns: what do we do? J Burn Care Rehabil 2004;25:71–5.
Results of a protocol of transfusion threshold and surgical technique on transfusion requirements in burn patients Michael S. O’Mara a,b, Fernando Hayetian c, Harvey Slater c, I. William Goldfarb c, Eric Tolchin c, Philip F. Caushaj c,* a
Shriner’s Hospital for Children of Northern California, Department of Burn Surgery, Sacramento, CA, USA b The University of California Davis Medical Center, Department of Surgery, Sacramento, CA, USA c The Western Pennsylvania Hospital, Department of Surgery, Burn Trauma Unit, Temple University School of Medicine Clinical Campus, 4800 Friendship Avenue, Pittsburgh, PA 15224, USA Accepted 6 January 2005
Abstract Introduction: Blood loss and high rates of transfusion in burn centers remains an area of ongoing concern. Blood use brings the risk of infection, adverse reaction, and immunosuppression. Methods: A protocol to reduce blood loss and blood use was implemented. Analysis included 3-year periods before and after institution of the protocol. All patients were transfused for a hemoglobin below 8.0 gm/dL. Results: Operations per admission did not change during the two time periods (0.78 in each). Overall units transfused per operation decreased from 1.56 0.06 to 1.25 0.14 units after instituting the protocol ( p < 0.05). Also, units transfused per admission decreased from 1.21 0.15 to 0.96 0.06 units of blood ( p < 0.05). This was noticed particularly in burns of less than 20% surface area, declining from 386 to 46 units after protocol institution, from 0.37 to 0.04 units per admission, and from 0.79 to 0.08 units per operation in this group of smallest burns. There was no change noted in the larger burns. Conclusions: This study suggests that a defined protocol of hemostasis, technique, and transfusion trigger should be implemented in the process of burn excision and grafting. This will help especially those patients with the smallest burns, essentially eliminating transfusion need in that group. # 2005 Elsevier Ltd and ISBI. All rights reserved. Keywords: Blood loss; Burn centers; Transfusion
1. Introduction Blood loss during excision and grafting of thermal injuries has long been of concern. Transfusion requirements have remained significant, with exsanguination sometimes life-threatening. Although replacement of blood loss is simple, the risk associated with transfusion is not insignificant. Blood products are related to a direct risk of infection [1], as well as an increased susceptibility to infection [2]. Also, transfusion has been implicated in the reduction of immunocompetence [3,4]. Consideration of these factors * Corresponding author. Tel.: +1 412 578 4024 (O); fax: +1 412 578 1434. E-mail address: [email protected] (P.F. Caushaj). 0305-4179/$30.00 # 2005 Elsevier Ltd and ISBI. All rights reserved. doi:10.1016/j.burns.2005.01.006
has led to an attempt to decrease transfusion requirements in burn patients, as well as the general surgical population [5]. A multitude of hemostatic techniques have been developed in recent years. Traditional methods of pressure and tourniquets are now being combined with chemotherapeutic agents [6]. Thrombin, epinephrine [7], and fibrin [8] are the most common modalities that have been used to minimize blood loss. This becomes vital as tangential excision is widely implemented, with the associated potential for increased blood loss. The traditional method of excision to fascia decreases operative blood losses, but is problematic for the extremities, as the sequelae of poor functional outcome are increased [9]. The benefits of minimizing blood loss become more apparent as strategies
M.S. O’Mara et al. / Burns 31 (2005) 558–561
involving a lowered threshold for transfusion are implemented, decreasing the number of patients who qualify for blood. Considering the advances in techniques to minimize blood loss and transfusion requirements, we retrospectively analyzed our institution’s transfusion practices prior to and following institution of a protocol to decrease intraoperative losses and de-liberalize transfusion threshold. In particular, we had noted that patients with smaller burns, those less than 20% total body surface area, rarely needed transfusion, and thus the risk seemed to have been eliminated. We sought to determine if a reduction in blood usage had occurred.
2. Material and methods Two 3-year time periods were analyzed, before and after implementation of our intraoperative protocol, to reduce blood loss. In the earlier period (1993–1995), methods of excision and grafting were more variable. In general, excision and harvesting was done initially, with coverage of sites with thrombin soaked and/or epinephrine soaked gauze. Pressure was applied to all sites for a minimum of 10– 15 min. Electrocautery was used to coagulate individual punctate hemorrhage, and pressure reapplied as needed. Hemostasis was achieved before grafting. Excision was done by a mixture of tangential excision with use of excision to fascia for extensive, deep burns of the thorax. In the later period (1997–1999), a protocol was followed, with modification based upon the attending surgeon’s clinical judgment. Extremities were excised tangentially under tourniquet control. The thorax was excised to fascia for burns of greater than 10% contiguous surface area involvement. Tangential excision was used for small, localized regions of the thorax (<5% contiguous area) and for female breast involvement. Whenever possible the nipple complex was preserved. Excision was done in an expeditious fashion, with two teams working in concert when more than one body region required excision. OmidermTM (Omiderm Ltd., commercially available) synthetic film sprayed with thrombin solution (GenTrac Inc., commercially available) was placed over all areas of excision and epinephrine-soaked gauze (1:1000 injectable, 1 mg, mixed in 1 L normal saline, American Regent Laboratories, commercially available) used to cover this. Pressure was then applied over the wound. While hemostasis was achieved, grafts were harvested. Over the split thickness donor site omidermTM, thrombin, and epinephrine soaked gauze were also applied. After 15 min of pressure, gauze and omidermTM were removed and wounds examined. Rare electrocautery was needed. Larger areas of diffuse capillary bleeding were re-covered until hemostasis was achieved. Donor skin was then meshed and grafted, and dressings applied. Postoperative bleeding was controlled with local electrocautery or direct suture ligation in the burn unit. The same two surgeons performed all procedures during both
559
time periods (IWG, HS). All procedures were performed at one hospital with patients all in the same burn unit (the Western Pennsylvania Hospital Burn Trauma Unit). During both periods, patients were evaluated for transfusion in the same manner. All patients were transfused for a hemoglobin below 8.0 gm/dL. Patients with a significant cardiac history and who were critically ill were transfused to a hemoglobin of 10.0 gm/dL. These patients were not considered separately. Patients were not given prophylactic transfusions in anticipation of blood loss if their hemoglobin concentration met these criteria. Intraoperative transfusion was reserved for patients who were hypotensive or tachycardic or were clinically judged to have excessive blood loss. Number of admissions, mean body surface area of burns, number of operations, area grafted per patient, units of blood transfused, units transfused per admission, and units transfused per operation were compared between the two time periods. All patients admitted with burn injuries were included in the evaluation. Criteria excluded only those patients admitted to the burn trauma unit who were not actually burn injuries (i.e. coumadin skin necrosis, toxic epidermal necrolysis). All thermal, chemical, and electrical burns were included whether or not operative interventions or transfusions were required. Patients refusing blood transfusion for religious reasons or otherwise were included. All operations were considered together, including those for re-grafting. Patient groups were further subdivided by burn size. The same evaluation was carried out, comparing small burns (0– 20%), moderate burns (20–40%), and large burns (>40%). Statistical analysis by two-tailed t-test was done using GraphPad InStat Version 3.01 for Windows 95/NT, GraphPad Software, San Diego, CA, USA (www.graphpad.com).
3. Results Total values were compared between the two time periods. Data are shown in Table 1. The later protocol group appeared to have smaller burns in all 11.2 14.5% versus 13.0 15.7% in the earlier group, and this was significant ( p = 0.01). Most of this difference appeared to be among the patients with smaller burns (see below). There was no change in number of operations per admission, 0.8 0.07 in the early group and 0.8 0.1 in the later, p = 1.0. No difference was noted in area grafted per patient, 995 cm2 in the early group and 1010 cm2 in the later, p = 0.9. There was a significant decrease in transfusions per admission and per operation. Transfusions per admission declined from 1.2 0.16 to 0.96 0.06 units ( p < 0.0001) and transfusions per operation from 1.6 0.14 to 1.3 0.14 units ( p < 0.0001). In the two groups of larger burns (20–40% and >40% surface area burns), there did not appear to be any change in transfusion requirements between the two time periods.
Total—all patients, all sizes of burn; small—patients with burns <20% surface area; mod—patients with burns 20–40% area; large—patient with burns >40% area. * p < 0.01, otherwise p > 0.05.
52 212 505 63.7 19.0 2685 2180 4.1 5.3 9.7 11.5 2.4 2.7 98 198 512 27.5 5.4 2060 1050 2.0 1.3 * 5.2 1.7 2.6 .67 111 215 542 28.7 5.5 1980 1120 1.3 1.2 * 4.9 1.6 2.5 .56 1129 567 46 6.6 5.3* 850 650 0.50 0.15* 0.04 0.01* 0.08 0.01* 1027 490 386 7.3 5.2* 805 605 0.48 0.10* 0.37 0.08* 0.79 0.12* 1271 981 1219 11.2 14.5* 1010 3820 0.78 0.11 0.96 0.06* 1.3 .14 * 1190 917 1433 13.0 15.7* 995 3870 0.78 0.07 1.2 .16 * 1.6 .14 * Admissions Operations Transfusion units Area of burn (%) Grafted area (cm2) Operation/admission Transfusion/admission Transfusion/operation
Large
Early Late
Mod
Early Late
Small
Early Early
Late Total
Table 1 Admission, operation, and transfusion data
44 216 661 64.5 18.8 2790 2040 4.9 5.9 15.0 14.9 3.1 5.3
M.S. O’Mara et al. / Burns 31 (2005) 558–561
Late
560
Burn size was comparable between the two time periods as well, as was the area grafted. The later group of moderate burns had more operations per admission, 1.3 1.2 versus 2.0 1.3 ( p < 0.001), but no change in transfusion need (Table 1). There was a trend in the large burns toward greater transfusions per admission and per operation, although this was not significant. The earlier group required 9.7 11.5 versus 15.0 14.9 units per admission in the later group, p = 0.051. And per operation, the earlier group was transfused 2.4 2.7 units and the later 3.1 5.3 units, p = 0.1 (Table 1). Consideration of the small burns revealed the greatest differences. In this segment, the later group had smaller overall burns, 6.6 5.3% versus 7.3 5.2% ( p = 0.009). Despite this, the later group had more surgical interventions, 0.50 0.15 versus 0.48 0.10 operations per admission in the earlier group, p = 0.001. This did not translate to a larger area of grafting. With more operations, smaller burns, and the hemostatic protocol, the later group had a significant drop in blood requirements. Grossly, this was a drop from 386 to 46 units over the respective time periods. The later period required fewer units per admission, 0.04 0.01 from 0.37 0.08, p < 0.0001. This protocol group also received fewer transfusions per operation, 0.08 0.01 units from 0.79 0.12 units earlier, p < 0.0001 (Table 1).
4. Discussion There has been a move across the spectrum of surgical specialties to reduce the necessity for blood transfusions. The benefits of decreased risk of blood borne infections, transfusion reactions, and immunosuppression have been documented [1–5]. In burn patients, it is as important to limit blood losses as it is to limit transfusion volume. Risk to cardiac function and hemodynamics of the patient are evident with exsanguination during burn surgery, and thus techniques to decrease these losses can both improve patient intraoperative risk as well as decreasing the need for postoperative transfusion. The evidence for risk of blood-born transmission and increased risk of infection with transfusion has been elucidated. An increased risk of bacterial infection exists postoperatively in the transfused patient. It is unlikely that these infections stem from infected units, but rather from host modulation that increases risk of clinical infection. The cause and effect is unclear, but the mere act of giving blood is associated with a likelihood of contamination [2]. HIV has been estimated to be a risk in up to 1 of every 200,000 units transfused, while nearly 1% of blood recipients are at risk of contracting hepatitis [10]. Screening of blood products is improving, but the risk remains. Immunosuppression with transfusion has been documented. Early recurrence of colorectal cancer is associated with perioperative transfusion, as well as increased likelihood of
M.S. O’Mara et al. / Burns 31 (2005) 558–561
postoperative infection. Leukocyte depletion in transfused units appears to be associated with improved outcomes, and may suggest the mechanism of immunomodulation. As of yet the clinical significance of this has not become clear, and results are conflicting as to the cause and effect of changed outcomes [10]. In studies of the critically ill and of surgical patients, hemoglobin levels as low as 7.0 gm/dL have shown no increase in mortality [11,12]. This is true in the majority of patients, but in patients who are at increased risk for cardiac mortality and who are critically ill, a hemoglobin of 9.0 gm/ dL has been suggested as being more prudent for adequate oxygenation and perfusion [12]. Prophylactic transfusions have not been shown to be of particular value, and may put a patient at unnecessarily increased risk [12,13]. Current practices at burn centers reveals a broad array of transfusion triggers and guidelines are being used [14]. No standard care exists despite current literature implying the efficacy of minimizing transfusions to critically ill patients. Overall we saw a decrease in transfusion use with the change in technique of hemostatic management. The majority of the decline in units of blood used appears to have been in the patients with the smallest burns. It is this group of patients where the benefit is also best seen. In larger burns, there would be only a modest benefit from decreasing the number of units transfused per patient or per operation, as the risk of transfusion is still present. In the smaller burns, though, there is a notable advantage in that transfusion was successfully eliminated in the majority of patients who previously would have been given blood. In the earlier time period, nearly 1 in 3 patients admitted received a transfusion, while after the protocol was instituted, this dropped to less than 1 patient transfused in every 25. The benefit to these patients, the largest group of burn patients, is considerable. We no longer type our patients with small burns for possible transfusion, and have attempted to decrease our frequency of evaluation of hemoglobin concentrations. The trend in the larger burns toward a greater transfusion requirement is concerning. This occurred despite no real difference in burn size. One contribution to this could be a trend toward more frequent, smaller excisions and grafting. We speculate that our confidence in better hemostasis allowing these patients to tolerate more has spurred us to be more aggressive in our excision and grafting of areas of burn we would have treated more conservatively in the past. This increased excision and grafting could benefit patients with a better functional outcome. Other modalities of decreasing blood loss may be more appropriate than our current methods, such as subeschar infusion of epinephrine and fibrin modalities. This could also decrease the need for excision to fascia in larger burns. This continues to be a process in evolution. Multiple means of decreasing blood loss and corresponding transfusion requirements have been used in the burn population. Techniques have been shown to individually
561
decrease losses. More paramount recently has been the identification of a need to move to a lower transfusion trigger. On combining these chemotherapeutic, surgical, and transfusion protocol methods, it is possible to bring about a coordinated diminution of transfusion. In order to optimize functional results, a certain level of bleeding must be accepted. By focusing efforts, we have shown it is possible to decrease overall use of blood in our burn unit, and in particular to eliminate transfusion requirements in a large segment of the burn population. Patients with larger burns can be more aggressively debrided and grafted toward better functional outcomes. Further studies comparing like burn sizes with direct measure of blood loss and hemoglobin change may be indicated, but we suggest that this is only necessary to confirm what we see, which is an improvement by decreasing blood need with the efforts of a burn team to minimize loss. This benefit is not limited to one specific protocol, but rather is a result of having a defined set of techniques and parameters in place. Such efforts decrease the multifactoral risk to the patient, and will undoubtedly improve outcomes.
References [1] Graves TA, Cioffi WG, Mason AD, McManus WF, Pruitt Jr BA. Relationship of transfusion and infection in a burn population. J Trauma 1989;29:948–52. [2] Triulzi DJ, Blumberg N, Heal JM. Association of transfusion with postoperative bacterial infection. Crit Rev Clin Lab Sci 1990;28:95– 107. [3] Persijn CG, Cohen B, Lansbergen Q, vanRood JJ. Retrospective and prospective studies on the effect of blood transfusions in renal transplantation in the Netherlands. Transplantation 1979;28:396–401. [4] Tartter PI. The association of perioperative blood transfusion with colorectal cancer recurrence. Ann Surg 1992;216:633–8. [5] Mann R, Heimbach DM, Engrav LH, Foy H. Changes in transfusion practices in burn patients. J Trauma 1994;37:220–2. [6] Sawada Y, Yotsuyanagi T. A technique of haemostasis of the extremities after debridement of burn wounds. Burns 1992;18:412–5. [7] Snelling CFT. Shaw K. The effect of topical epinephrine hydrochloride in saline on blood loss following tangential excision of burn wounds. Plast Reconstr Surg 1983;72:830–4. [8] Ihara N, Suzuki K, Tanaka H, Nakamura Y, Tanabe J, Fukuda A, et al. Application of fibrin glue to burns. Burns 1984;10:396–404. [9] Achauer BM, Hernandez J, Parker A. Burn excision with intraoperative vasopressin. J Burn Care Rehabil 1989;10:375–8. [10] Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion medicine. NEJM 1999;340:438–47. [11] Carson JL, Duff A, Jerlin JA, Lawrence VA, Poses RM, Huber EC, et al. Perioperative blood transfusion and postoperative mortality. JAMA 1998;279:199–205. [12] Hebert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. NEJM 1999;340:409–17. [13] Valeri CR, Crowley JP, Loscalzo J. The red cell transfusion trigger: has a sin of commission now become a sin of omission? Transfusion 1998;38:602–10. [14] Palmieri TL, Greenhalgh DG. Blood transfusion in burns: what do we do? J Burn Care Rehabil 2004;25:71–5.